The thermal contribution to photoactivation in A2 - Ala

Visual Neuroscience (2003), 20, 411–419. Printed in the USA.
Copyright © 2003 Cambridge University Press 0952-5238003 $16.00
DOI: 10.10170S0952523803204065
The thermal contribution to photoactivation in A2 visual pigments
studied by temperature effects on spectral properties
PETRI ALA-LAURILA,1 RAULI-JAN ALBERT,1 PIA SAARINEN,2 ARI KOSKELAINEN,1
and KRISTIAN DONNER 2
1
2
Laboratory of Biomedical Engineering, Helsinki University of Technology, FIN-02015 HUT, Finland
Department of Biosciences, Division of Animal Physiology, FIN-00014 University of Helsinki, Finland
(Received April 23, 2003; Accepted July 1, 2003)
Abstract
Effects of temperature on the spectral properties of visual pigments were measured in the physiological range
(5–288C) in photoreceptor cells of bullfrog (Rana catesbeiana) and crucian carp (Carassius carassius). Absorbance
spectra recorded by microspectrophotometry (MSP) in single cells and sensitivity spectra recorded by
electroretinography (ERG) across the isolated retina were combined to yield accurate composite spectra from ca.
400 nm to 800 nm. The four photoreceptor types selected for study allowed three comparisons illuminating the
properties of pigments using the dehydroretinal (A2) chromophore: (1) the two members of an A10A2 pigment pair
with the same opsin (porphyropsin vs. rhodopsin in bullfrog “red” rods); (2) two A2 pigments with similar spectra
(porphyropsin rods of bullfrog and crucian carp); and (3) two A2 pigments with different spectra (rods vs.
long-wavelength-sensitive (L-) cones of crucian carp). Qualitatively, the temperature effects on A2 pigments were
similar to those described previously for the A1 pigment of toad “red” rods. Warming caused an increase in relative
sensitivities at very long wavelengths but additionally a small shift of lmax toward shorter wavelengths. The former
effect was used for estimating the minimum energy required for photoactivation ~Ea ) of the pigment. Bullfrog rod
opsin with A2 chromophore had Ea 5 44.2 6 0.9 kcal0mol, significantly lower (one-tailed P , 0.05) than the value
Ea 5 46.5 6 0.8 kcal0mol for the same opsin coupled to A1. The A2 rod pigment of crucian carp had Ea 5 42.3 6
0.6 kcal0mol, which is significantly higher (one-tailed P , 0.01) than that of the L-cones in the same retina
~Ea 5 38.3 6 0.4 kcal0mol), whereas the difference compared with the bullfrog A2 rod pigment is not statistically
significant (two-tailed P 5 0.13). No strict connection between lmax and Ea appears to exist among A2 pigments
any more than among A1 pigments. Still, the A1 r A2 chromophore substitution in bullfrog opsin causes three
changes correlated as originally hypothesized by Barlow (1957): a red-shift of lmax , a decrease in Ea , and an
increase in thermal noise.
Keywords: Photoreceptor, Spectral sensitivity, Activation energy, Porphyropsin, Retinal
encountered in fishes and amphibians as seasonal or developmental changes (Dartnall et al., 1961; Bridges, 1964; Reuter, 1969;
Allen & McFarland, 1973; McFarland & Allen, 1977) has two
physiologically relevant effects. First, it red-shifts the wavelength
of peak absorption (lmax ) of the visual pigment by typically more
than 20 nm and broadens the absorbance spectrum (Dartnall &
Lythgoe, 1965; Bridges, 1967; Bridges & Yoshikami, 1970; for
review see Bridges, 1972; about spectral broadening, see Makino
et al., 1999). Even partial A1 r A2 substitution strongly enhances
sensitivity to long wavelengths. Second, the switch decreases the
thermal stability of the pigment molecule. Porphyropsin in extract
decays more rapidly than rhodopsin (Dartnall, 1955; Bridges,
1956, 1967; Williams & Milby, 1968), and this probably reflects
the same underlying difference as the at least ten-fold higher rate
of photon-like “dark” events (spontaneous pigment activations)
observed in the receptor current of porphyropsin rods compared
with rhodopsin rods of the bullfrog (Donner et al., 1990). Thus, the
chromophore is an important determinant of the two most impor-
Introduction
Visual pigments form a large family of 7-TM receptors, where the
protein (opsin) is covalently bound to a light-absorbing prosthetic
group, the chromophore. Vertebrate pigments use either of two
forms of chromophore: 11-cis-retinal, denoted A1, or 11-cis-3,4dehydroretinal, denoted A2. A1 pigments are commonly called
rhodopsins, and A2 pigments porphyropsins. In birds and mammals, only A1 has been found, whereas many fish and amphibian
species can also utilize A2 or a mixture of both (Dartnall &
Lythgoe, 1965; Bridges, 1972). Compared with A1, the A2 chromophore has an additional double bond in the b-ionone ring, a
difference which a priori is expected to lower the energy barrier
for isomerization (Bridges, 1956, 1967). The switch from A1 to A2
Address correspondence and reprint requests to: Petri Ala-Laurila,
Laboratory of Biomedical Engineering, P.O. Box 2200, Helsinki University
of Technology, FIN-02015 HUT, Finland. E-mail: [email protected]
411
412
tant functional properties of the visual pigment: spectral absorbance, which defines the visible range of electromagnetic radiation,
and the rate of thermal activation, which sets an ultimate limit to
visual sensitivity in darkness (Autrum, 1943; Barlow, 1956; Ashmore & Falk, 1977; Aho et al., 1988).
The absorbance spectra of visual pigments show subtle changes
with temperature not only in extracts under extreme cooling (St.
George, 1952; Yoshizawa & Horiuchi, 1969; Yoshizawa, 1972),
but even in vivo. One effect is a small shift of lmax , which at least
in the A1 “red” rods of toad is clearly measurable even in the
physiological temperature range (Ala-Laurila et al., 2002). Most
interesting, however, is the relative increase in long-wavelength
sensitivity with rising temperature (de Vries, 1948), which according to theory originally put forward by Stiles (1948) allows
estimation of the minimum energy needed for photoactivation ~Ea )
of the pigment molecule (Denton & Pirenne, 1954; Srebro, 1966;
Koskelainen et al., 2000). Ala-Laurila et al. (2002) showed that the
temperature effect on the A1 pigment of toad “red” rods is in
agreement with Stiles’ theory as developed by Lewis (1955) and
estimated Ea on this basis.
Here we use the same methods to study four other visual
pigments in situ in photoreceptor cells, selected to allow three
comparisons that elucidate the properties imparted by the A2
chromophore. The first comparison is between an A2 pigment and
its A1 counterpart with the same opsin (bullfrog porphyropsin vs.
rhodopsin rods). The second comparison is between two A2 pigments with similar absorbance spectra (porphyropsin rods of bullfrog vs. crucian carp). The third comparison is between two A2
pigments with different spectra (rods vs. L-cones of crucian carp).
Thus, in the latter two comparisons the opsins are varying while
the chromophore (A2) is the same.
The results are consistent with our earlier conclusion for A1
pigments that no unique and inevitable physical connection exists
between lmax and Ea (Koskelainen et al., 2000; Ala-Laurila et al.,
2002). However, for the one opsin where all parameters are now
available, the visual pigment of bullfrog “red” rods, the changes in
lmax , Ea , and thermal noise due to the chromophore substitution
per se correlate qualitatively as predicted by Barlow (1957). The
spectral red-shift achieved by switching from A1 to A2 is indeed
associated with lowered activation energy and increased thermal
noise.
Methods
Animals—preparations and recording
Crucian carp (Carassius carassius, body length ca. 7–15 cm) were
caught in the autumn in a pond in SW Finland, kept in aquaria at
158C on a 12-h light012-h dark regime and fed with standard food
for aquarium fish. Medium-sized bullfrogs (Rana catesbeiana,
body length ca. 10 cm) were obtained from Carolina Biological
Supply Co. (Burlington, NC). They were kept in either of two
conditions, favoring or suppressing the development of significant
proportions of A2 pigment in dorsal rods (Reuter et al., 1971).
Porphyropsin-favoring conditions implied a 12-h light012-h dark
light regime at 158C in basins with a white floor, lit from above by
fluorescent tubes (Tsin & Beatty, 1980; Donner et al., 1990). The
second, “rhodopsin” group was kept in natural daylight (light–dark
cycle varying from 12-h012-h to 18-h06-h during the summer
half-year in Helsinki) at 208C in grey plastic basins with covers
strongly attenuating the illumination. The pieces of retina used for
A2 experiments were taken from a crescent-shaped field near the
P. Ala-Laurila et al.
dorsal rim of the retina, where the A2 proportion is maximal
containing up to 81–89% porphyropsin (Reuter et al., 1971; Donner et al., 1990). The pieces of retina for the A1 experiments were
taken from the central retina near the optic disc, least likely to
contain A2.
This work is based on two types of recordings. The spectral
absorbance of visual pigments in situ was studied by microspectrophotometry (MSP) from the outer segments of isolated rods or
cones. The spectral sensitivities of the corresponding photoreceptor types were determined electrophysiologically by ERG recording of mass photoresponses across the isolated, aspartate-superfused
retina. MSP is superior for spectral characterization in the main
absorbance band of the visual pigment, whereas electrophysiology
is superior in the long-wavelength range where pigment absorbance is low (cf. Govardovskii et al., 2000).
All rod recordings were performed at two “physiological”
temperatures (8.5 6 0.58C, “cold”, C and 28.5 6 0.58C, “warm”,
W) from the same sample of isolated cells (MSP recordings) or the
same retina (ERG experiments). The experiments on bullfrog
porphyropsin rods further comprised MSP recordings at room
temperature (218C) used for determination of the A10A2 ratio in
each experiment by fitting of room-temperature spectral templates
(Govardovskii et al., 2000). In crucian carp L-cones, ERG was
recorded at three different temperatures: 5.0 6 0.58C, 15.0 6
0.58C, and 25.0 6 0.58C, whereas MSP was done only at room
temperature (218C).
Our conclusions are based on temperature-dependent changes
in relative sensitivities in the long-wavelength range, that is, in the
range studied by ERG. One requirement for this, however, is the
availability of reliable measurements over an essentially temperatureinvariant spectral domain to allow correct normalization of “warm”
and “cold” data sets relative to each other. The MSP recordings
provide such a reference, to which the long-wavelength ERG data
could be anchored. The ERG spectra are much less reliable at
shorter wavelengths due to several potential artifacts (Ala-Laurila
et al., 2002).
For a detailed description of most aspects of the Methods, the
reader is referred to Donner et al. (1988), Koskelainen et al. (1994,
2000), Govardovskii et al. (2000), and especially Ala-Laurila et al.
(2002). Here we focus on a few special procedures required in the
present work (1) for separating cone from rod responses in the
crucian carp ERG; (2) for extracting pure A2 (porphyropsin)
spectra from MSP recordings in bullfrog rods containing A10A2
mixtures; and (3) for analysing the ERG data from crucian carp
cones, where MSP data at the correct temperatures are not available.
Isolation of L-cone sensitivities from the ERG mass
potential of the crucian carp retina
Even when synaptic transmission has been blocked by aspartate,
the ERG signal recorded across the isolated retina is the sum of
several (ohmic voltage) components, notably those due to photocurrents of more than one photoreceptor types and Müller cell
(glial) currents elicited by changes in extracellular potassium near
the photoreceptors. The glial contribution does not, however,
distort spectral sensitivities as measured here (see Ala-Laurila
et al., 2002).
Cone photoresponses were isolated from rod responses by a
double-flash technique (Fig. 1), utilizing the fact that flash sensitivity is higher and response recovery slower in rods than in cones.
A first, rod-saturating conditioning flash was followed after a fixed
delay by a second, cone-stimulating test flash. The test flash was
Temperature effects on porphyropsin spectra
413
Extraction of the porphyropsin spectrum from mixed
A1/A2 spectra of bullfrog
Fig. 1. The double-flash technique for isolating cone responses in the
aspartate-isolated mass photoresponse recorded across the isolated retina
of crucian carp. A rod-saturating blue (494 nm) conditioning flash at time
t 5 0 elicited a mixed response of rods and cones. When the cones had
recovered but the rod response remained saturated, this was followed by a
dim cone-stimulating test flash (at t 5 2 s in the figure), the wavelength of
which was varied. The trace is a single recording. Homogeneous 20-ms
flashes, sampling frequency 200 Hz. The Inset shows a family of averaged
(2– 4) responses to flashes of increasing intensity at our reference wavelength for L-cone recordings, 702 nm (cone responses were low-pass
filtered with 100-Hz cutoff ). The responses were recorded in sequence just
after the example trace shown in the main panel and with a similar
protocol. Response families recorded at the reference wavelength were
used for constructing the L-cone I–R functions needed for determining
spectral sensitivities from small responses to test wavelengths (Koskelainen et al., 2000; Ala-Laurila et al., 2002). The temperature in the
recordings shown was 158C.
timed to occur when the cones had recovered, but while the rods
still remained firmly saturated (this was assessed as described by
Koskelainen et al., 1994). The required delay was 1–3 s (2 s in
Fig. 1) depending on the temperature of the experiment. Michaelistype saturation behavior was taken as indication that only one cone
type was involved at the reference wavelength (702 nm). On these
criteria, the response family to 702-nm flashes of increasing
intensity shown in the inset of Fig. 1 is thought to represent the
characteristics of essentially dark-adapted L-cones.
With test flashes of shorter wavelengths, two major artifacts
may intrude. First, responses may contain contributions from other
cone types, especially from the green-sensitive (M-) cones. According to our present MSP measurements, crucian carp retina
contains four cone types, similar to those of the closely related
goldfish, Carassius auratus (Marks, 1965; Stell & Hárosi, 1976;
Palacios et al., 1998). Thus, we found (1) L-cones, lmax 5 619 nm,
(2) M-cones, lmax 5 531 nm, (3) S-cones, lmax 5 456 nm, and (4)
UV-cones, lmax ca. 350–370 nm (all estimates obtained by fitting
the A2 template of Govardovskii et al. (2000)). Second, as the
retina was illuminated from the photoreceptor side, there could be
substantial selective screening by the rod pigment, which has
lmax 5 526 nm at room temperature. As we wished to deal with
undistorted L-cone sensitivities, the study was limited to l $
680 nm, where the sensitivity of the M-cone and the absorbance of
the rod visual pigment both have dropped by more than 2 log units
from peak.
Bullfrog “porphyropsin” rods in fact always have a mixture of A2
and A1 chromophores (Donner et al., 1990). The lmax value of
the A1 pigment has been measured with considerable accuracy
(Reuter et al., 1971; Donner et al., 1990; Govardovskii et al.,
2000). That of its A2 counterpart can be roughly defined by the
general relation between rhodopsins and corresponding porphyropsins (Dartnall & Lythgoe, 1965; Hárosi, 1994), but greater
precision was attained as follows. The good general templates
available for A1 and A2 pigment spectra (Govardovskii et al.,
2000) made it possible to determine what proportion of “mixed”
absorbances was due to each by fitting the recorded spectrum with
a linear combination of the two templates. Thus, we could purify
the A2 absorbance component from spectra “contaminated” by A1
chromophore. The lower molar extinction coefficient of the A2
pigment need not be taken into account as long as the analysis is
restricted to A1 and A2 absorbance components (which is how it
will be used here unless otherwise stated). In terms of numbers of
visual-pigment molecules underlying the absorbances, the proportions would of course be different, and this would be relevant, for
example, if we wished to consider the amount of pigment-related
thermal noise associated with a certain spectrum.
With respect to combining MSP and electrophysiology for
studying rods with mixed chromophore content, two further points
need to be emphasized. First, there is no significant difference in
the quantum efficiency of bleaching between A1 and A2 pigments
(see e.g. Dartnall, 1972). This means that the probability for an
absorbed photon to isomerize the chromophore and trigger phototransduction is equal in the two. Second, response univariance
holds at least for bullfrog rods, as studied here (Firsov et al., 1994).
This means that the size or shape of the quantal response does not
depend on whether it is initiated by a rhodopsin or porphyropsin
molecule. The two variables by which we characterize spectral
properties, outer-segment absorbance and flash sensitivity, are
therefore directly proportional.
After each ERG experiment on bullfrog “porphyropsin” rods,
pieces of the preparation from which the recordings had been made
were transferred in darkness from the specimen holder to a Ringer
dish and samples for MSP were prepared. MSP spectra were then
recorded at room temperature (ca. 218C), in each case from ca. 20
rods sampled from the retinal field just studied by ERG or from its
immediate neighborhood. Thus, each ERG experiment is accompanied by an MSP spectrum recorded at the same temperature as
those on which the Govardovskii et al. (2000) templates are based,
the purpose here being accurate determination of the A10A2 ratio.
In part of the experiments, the ERG recordings were supplemented
by two further MSP spectra recorded at the experimental temperatures of the ERG (ca. 8.58C and at 28.58C). Additional
“porphyropsin-rod” MSP spectra at these temperatures were collected from samples of dorsal retina that had not been used in ERG
experiments. In two control experiments, MSP spectra were recorded at all three temperatures from the same sample of 12
individual cells to assess to what extent variability of the A10A2
ratio may cause spurious “temperature effects” in the main body of
experiments, where different samples of rods were recorded at
each temperature. The control experiments showed this possible
source of error to be insignificant.
Besides knowledge of the A10A2 ratio, the second prerequisite
for extracting pure A2 spectra from mixed spectra at different
temperatures is the availability of pure A1 spectra recorded at the
414
P. Ala-Laurila et al.
Fig. 2. The method for estimating the proportion of absorbance due to
A2 visual pigment in bullfrog “porphyropsin” rods. Main panel: The
black circles show logarithmic spectral sensitivities vs. wavenumber
(10l) measured in a single ERG experiment at 28.58C on a piece of
dorsal bullfrog retina. The black curves show linear combinations of A1
and A2 templates (Govardovskii et al., 2000) corresponding to different
A10A2 absorbance proportions (0–100% in 20% steps). The dashed red
line is the best-fitting template (corresponding to 71% A2 absorbance),
carried over from the inset. The green trace is the synthetic MSP
spectrum for 71% A2 at 28.58C (see Text). Inset: The same templates
(black curves) on linear scales, compared with an MSP spectrum
averaged from 20 isolated cells taken from the same piece of retina and
recorded at room temperature (218C) (violet circles). The red line
shows the best-fitting Govardovskii et al. (2000) template (least-square
criterion over the interval 470–780 nm), corresponding to 71% A2
absorbance.
same temperatures. These were obtained from mid-ventral rods of
“rhodopsin” retinas, containing no measurable amount of A2 chromophore as judged by template fitting. Assuming as small an
admixture as 1–2% of A2 degraded the fits perceptibly compared
with the pure A1 template. The A1 spectra were averaged across all
experiments at each of the experimental temperatures.
Linear combinations of A1 and A2 templates, with the A2
percentage going from 0% to 100% in 20% steps, are shown in
Fig. 2 (smooth continuous lines). The main panel presents the
spectra on logarithmic ordinates against wavenumber, the Inset on
linear scales. Especially the logarithmic format makes it obvious
how strongly spectral sensitivities at long wavelengths are dominated by A2 even at quite low percentages of this chromophore
(e.g. with only 20% A2, almost 90% of the sensitivity at wavelengths $ 650 nm is due to porphyropsin). The open black circles
in the main panel give an example of a spectrum recorded in a
single ERG experiment from the porphyropsin field of the retina at
28.58C
The violet circles in the Inset shows the room-temperature MSP
spectrum averaged from 20 rods from the same piece of retina that
yielded the ERG data in the main panel. It lies between the
templates for 60% and 80% A2, and the best fit corresponds to
71% A2 (red continuous line). The A10A2 absorbance proportion
in each ERG experiment was determined from room-temperature
MSP in this way. For each experiment and temperature, we then
recovered a pure A2 spectrum by subtracting the relevant proportion of the average A1 spectrum for that temperature from the
mixed spectrum. The pure A2 spectra were averaged across experiments, separately for 8.5 and 28.58C.
A high-quality “mixture” MSP spectrum was resynthesized for
each experiment and temperature by summing the correct proportions of the averaged A2 and A1 spectra, and the relevant ERG
data were anchored to this to yield a composite (MSP 1 ERG)
spectrum. In the main panel of Fig. 2, the green trace is the
synthetic MSP spectrum for 71% A2 at 28.58C. The circles marking the ERG data have been joined to the MSP spectrum by
least-square fitting based on the three ERG points (601, 615, and
621 nm) where the two data sets overlap. Before anchoring of the
ERG data, the noisy long-wave tail of the MSP spectrum was
smoothed by fitting of a second-order polynomial (cf. Ala-Laurila
et al., 2002). Thus a composite “mixture” spectrum was obtained.
For comparison, a linear combination of the relevant Govardovskii
et al. (2000) A1 and A2 templates (lmax,A1 5 501.6 nm) for 71%
A2 is shown as a dashed red line in the main panel. Weighted
subtraction of the A1 composite spectrum from the mixture composite spectrum gave a pure A2 composite spectrum at each
temperature (see Fig. 3 below). Before subtraction, the A1 spectra
were smoothed by fitting of a second-order polynomial to the
long-wavelength tail (594–777 nm).
Analysis of the ERG data from crucian carp cones
Obtaining good enough MSP records from crucian carp cones to
discern possible temperature effects near lmax proved difficult. In
Fig. 3. Averaged composite spectra (log-normalized absorbance and lognormalized spectral sensitivity) of bullfrog rhodopsin and porphyropsin at
8.58C (“cold”) and at 28.58C (“warm”). MSP data are shown by lines: blue
dashed line (cold) and red continuous line (warm). ERG data are shown by
blue and red symbols corresponding to “cold” and “warm” data, respectively: blue crosses and red squares (rhodopsin rods); blue plus signs and
red circles (porphyropsin rods). The ERG data have been anchored with the
least-square method to the corresponding MSP spectrum at the three
shortest wavelengths, where two are seen to overlap (rhodopsin rods:
594– 621 nm, porphyropsin rods: 615– 638 nm). The porphyropsin spectra
shown have been purified from mixed A10A2 spectra (see Methods).
Temperature effects on porphyropsin spectra
this case, therefore, we contented ourselves with MSP recordings
at room temperature to get an accurate value for lmax . The rest of
the analysis was based on ERG data alone. ERG recordings were
conducted at three temperatures (5, 15, and 258C) over the longwavelength domain where spectral distortion from M-cone responses and screening by rods can be neglected ($ 680 nm, see
above). From the original spectra, difference spectra (15–58C,
25–158C, and 25–58C) were formed without any previous normalizing assumptions (Ala-Laurila et al., 2002). Normalization is here
deferred to a posteriori definition of a zero-difference baseline,
which must be based on a wavelength domain where the relation
between “warm” and “cold” sensitivities is supposedly temperature invariant. In fact, the experimental difference spectra did
display a reasonably horizontal plateau (indicating constant warm0
cold sensitivity ratio) abutting to the warming-induced rise at long
wavelengths (see Fig. 5). The normalization of the difference
spectra was therefore done by least-square fitting of the “plateau”
data points to a zero line.
Results
Changing temperature affected the absorbance or sensitivity spectra of the A2 pigments (and, of course, the bullfrog A1 pigment as
well) in two ways, consistent with results previously obtained in
the A1 rods of the toads Bufo marinus and Bufo bufo (Ala-Laurila
et al., 2002). First, there was a small shift in the wavelength
of maximum absorption lmax , averaging ca. 20.6 nm per 108C
temperature rise in the two A2 rod pigments studied here, bullfrog
and crucian carp (data not shown). For these comparisons, the lmax
of each spectrum was determined by a polynomial fit to the data
around peak (to absorbance values A $ 0.9, see Ala-Laurila et al.,
2002). The observed shift in lmax is roughly consistent with
interpolation to physiological temperatures of the shift observed
upon cooling a carp porphyropsin extract from 115 to 21908C,
where spectra published by Yoshizawa and Horiuchi (1969) and
Yoshizawa (1972) indicate 20.4 nm per 108C. A detailed consideration of the similar effect on lmax of rhodopsins can be found in
Ala-Laurila et al. (2002).
Second, there was an increase in relative sensitivities to long
wavelengths, starting from some limit l 0 . l max and getting larger
towards longer wavelengths. It is this effect that concerns us here,
because it can be used for estimation of the minimum energy for
photoactivation ~Ea ) of the pigments (Stiles, 1948; Lewis, 1955;
Srebro, 1966; Koskelainen et al., 2000; Ala-Laurila et al., 2002).
To measure the relative changes in long-wavelength sensitivities
with sufficient accuracy, we used spectrophotometry and electrophysiology in the respective domains where each is superior and
glued them together to form composite MSP 1 ERG spectra (see
Methods, especially Fig. 2).
MSP spectra were collected from 98 (8.58C) and 84 (28.58C)
rhodopsin rods (3 different animals) and from 97 (8.58C) and 57
(28.58C) porphyropsin rods from five different animals. ERG
spectra were recorded from the ventral (rhodopsin) field in five
retinas at 8.58C and four retinas at 28.58C and from the dorsal
(porphyropsin) field in eight (8.58C) and ten (28.58C) retinas.
Fig. 3 shows composite “cold” (8.58C) and “warm” (28.58C)
spectra for bullfrog rhodopsin and porphyropsin rods, plotted on
logarithmic ordinates against wavenumber (10l). The lines trace
the MSP data (dashed: cold, continuous: warm). For the A1
pigment, the MSP data are reproduced as such, whereas the A2
curves are conceptually purified spectra, where the contribution of
some measured proportion of A1 present in the porphyropsin rods
415
studied has been subtracted (see Methods). The symbols give the
spectral sensitivities measured by ERG. Oblique crosses and squares
give the averaged A1 data for the “cold” and “warm” conditions,
respectively, upright crosses and circles give the averaged, purified
A2 data for the two temperatures. The main point to observe is
the divergence of “warm” and “cold” spectral sensitivities at long
wavelengths.
Fig. 4A shows this interesting range on expanded scales.
This format makes it easier to appreciate quantitatively the
differences in the long-wavelength range and it also does justice
to the fairly high precision of the measurements (note the SEM
bars, which would have been obscured by the size of the
symbols if drawn in Fig. 3). Panels B and C display the data for
the two crucian carp pigments (B: 526-nm rods, C: 619-nm
cones). Both these cell classes have A2 pigments with no measurable contribution from A1, and it is immediately evident that
the temperature effects are smaller, emerging only at longer
wavelengths compared especially with the bullfrog A1 pigment.
On Stiles’ (1948) theory, each warm–cold sensitivity difference
can be used for an independent point estimate of the energy of
photoactivation Ea (Denton & Pirenne, 1954; Srebro, 1966).
Calculating Ea for each of the four pigments from the data
shown in Fig. 4, with details as given by Koskelainen et al.
(2000) and Ala-Laurila et al. (2002), we obtained the following values: bullfrog rhodopsin 46.5 6 0.8 kcal0mol, bullfrog
porphyropsin 44.2 6 0.9 kcal0mol, crucian carp porphyropsin
42.3 6 0.6 kcal0mol, and crucian carp L-cone pigment 38.3 6
0.4 kcal0mol. The differences were statistically assessed by
means of Student’s t-test. When considering an A2–A1 pair
with the same opsin, or two A2 pigments with significantly
different lmax , theory as well as earlier results (Barlow, 1957;
Koskelainen et al., 2000) warranted that the zero hypothesis
(“no difference”) be confronted with one-tailed counter-hypotheses
of the form: in case differences exist, the A2 pigment has lower
Ea than its A1 pair, and the more red-sensitive pigment has
lower Ea than the less red-sensitive one (with the same chromophore). The bullfrog A2–A1 difference is statistically significant
~P , 0.05), as is that of crucian-carp L-cone versus rod pigment ~P , 0.01). However, comparing the spectrally similar
bullfrog and crucian-carp rod A2 pigments required two-tailed
testing, and in this case the zero hypothesis could not be rejected ~P 5 0.13).
In Fig. 4, the data for all four pigments have been fitted with
straight lines by weighted linear regression on log S 2 10l scales
(the weighting factor for each point being 10SEM 2 ). The fits of the
straight lines are rather acceptable overall and in qualitative agreement with Stiles’ (1948) theory. Moreover, the regression coefficients (slopes) are on the order of magnitude predicted by the
theory (hc0kT ln 10),* although somewhat shallower. The values
are as follows: bullfrog rhodopsin, 1.64 3 1025 m (8.58C) and
1.55 3 1025 m (28.58C); bullfrog porphyropsin, 1.66 3 1025 m
(8.58C) and 1.57 3 1025 m (28.58C); crucian carp porphyropsin,
1.57 3 1025 m (8.58C) and 1.45 3 1025 m (28.58C); and crucian
carp L-cone pigment, 1.99 3 1025 m (5.08C) and 1.77 3 1025 m
(258C). These slopes lie between 70% and 89% of the Stiles
prediction. A similar degree of agreement (76% of the theoretical
value) was found by Ala-Laurila et al. (2002) in rhodopsin rods of
Bufo marinus. Obviously, the theory further predicts that the ratio
*h 5 Planck constant, c 5 speed of light, k 5 Boltzmann constant, T 5
absolute temperature.
416
P. Ala-Laurila et al.
of slopes KW and KC at two temperatures TW and TC should be
KC 0KW 5 TW 0TC . This factor is (301.7 K)0(281.7 K) 5 1.071 for
our experimental temperatures 8.58C and 28.58C and 1.072 for the
temperatures 5.08C and 25.08C used in the L-cone experiments.
The experimental ratios ~KC 0KW ) obtained were 1.062 (bullfrog
rhodopsin), 1.054 (bullfrog porphyropsin), 1.079 (crucian carp
rhodopsin), and 1.123 (crucian carp L-cone pigment).
Fig. 5 summarizes the data in terms of the warm–cold differences in logarithmic relative absorbances or logarithmic relative
sensitivities, plotted against 10l. We shall term these functions
D-spectra. They have been normalized with respect to a baseline
that represents the absence of temperature effects (see Methods).
Upward deviations from the zero line indicate relative increases in
“warm” sensitivities.
According to Stiles’ (1948) theory and its refinement by Lewis
(1955), D-spectra should follow the baseline exactly up to the
long-wavelength domain (beyond l0 ), where the relative increase
in “warm” sensitivities sets in as a monotonical rise. In Lewis’
formulation, the rise from baseline would start gently and gradually steepen (depending on the parameter m of the model) to reach
finally the constant slope predicted by Stiles. In the range where
the warm–cold differences have grown larger than 0.1 log units,
the Lewis prediction converges with that of Stiles, envisaging
straight lines of slope
KD 5
Fig. 4. The long-wavelength tails of ERG sensitivity spectra after anchoring to MSP. (A) Bullfrog rhodopsin and porphyropsin rods, (B) crucian
carp rods, and (C) crucian carp L-cones. Spectral sensitivities at “cold”
temperature (8.58C in rod recordings, 5.08C in crucian carp cone recordings) and at “warm” temperature (28.58C in rod recordings, 25.08C in
crucian carp cone recordings) are shown by blue and red symbols, respectively. The error bars are SEMs. The data points are means of recordings
from 4–10 retinas depending on the photoreceptor type and the temperature. The straight lines are fitted by weighted regression (weighting coefficient of each data point 5 10SEM 2 ). The wavelength intervals of the
regression analysis, the regression slopes, and the coefficients of determination ~r 2 ) are as follows: bullfrog rhodopsin rods (653 nm–777 nm),
1.64 3 1025 m (8.58C), r 2 5 0.999 (8.58C), 1.55 3 1025 m (28.58C), r 2 5
0.999 (28.58C); bullfrog porphyropsin rods (701 nm–803 nm), 1.66 3
1025 m (8.58C), r 2 5 0.998 (8.58C), 1.57 3 1025 m (28.58C), r 2 5 0.9995
(28.58C); crucian carp red rods (660 nm–777 nm), 1.57 3 1025 m (8.58C),
r 2 5 0.998 (8.58C), 1.45 3 1025 m (28.58C), r 2 5 0.999 (28.58C); and
crucian carp L-cones (750 nm–803 nm), 1.99 3 1025 m (5.08C), r 2 5 0.997
(5.08C), 1.77 3 1025 m, r 2 5 0.996 (25.08C).
hc~TC 2 TW )
kTW TC ln 10
,
(1)
Fig. 5. Warm–cold difference spectra (D log S 5 log Sw 2 log Sc ), showing
the temperature effect on log relative spectral sensitivity and absorbance.
MSP data are shown by small, filled symbols and ERG data are shown by
larger open symbols (red: bullfrog rhodopsin rods, green: bullfrog porphyropsin rods, blue: crucian carp rods and pink: crucian carp L-cones). The
dashed line shows the zero baseline determined by normalization of cold
and warm MSP spectra to the same peak value. The error bars are SEMs.
The straight black lines are constrained to have the slopes predicted by
Stiles’ model (21.47{1026 m and 21.51{1026 m corresponding to the
temperatures of rod and cone recordings, respectively) and have been
positioned by a least-square criterion to fit the long-wavelength points for
which D log S $ 0.1 (for the crucian carp cones, both the data points that
rise above baseline were accepted, implying loosening the criterion to
D log S $ 0.06). The points of intersection of the zero line and “Stiles”
fits give the following graphical estimates of the activation energy:
45.2 kcal0mol (bullfrog rhodopsin rods), 43.0 kcal0mol (bullfrog porphyropsin rods), 41.9 kcal0mol (crucian carp rods), and 38.3 kcal0mol (crucian
carp L-cones).
Temperature effects on porphyropsin spectra
which for TC 5 281.7 K and TW 5 301.7 K gives 21.47{1026 m.
Straight lines with this slope have therefore been positioned for
best fit to those data points that lie more than 0.1 log units above
baseline. (A somewhat looser criterion was applied to crucian carp
cones, see figure legend.) The lines are not inconsistent with the
data, although the data as such would not, of course, suffice to
determine the slopes with high precision. According to Stiles’
original formulation, these lines intersect with the baseline at l0 ,
thus defining the energy of photoactivation Ea 5 hc0l0 . The Ea
values derived in this way from Fig. 5 are 45.2 kcal0mol for
bullfrog rhodopsin, 43.0 kcal0mol for bullfrog porphyropsin, 41.9
kcal0mol for crucian carp porphyropsin, and 38.3 kcal0mol for
crucian carp L-cone pigment. These values are in fair agreement
with the mean point estimates derived above (46.5, 44.2, 42.3, and
38.3 kcal0mol), but we would like to emphasize the primacy of the
latter. While both methods are sensitive to limitations of the theory
itself, the more simply “empiricist” point-by-point estimation is
somewhat less vulnerable (see Ala-Laurila et al., 2002). One
“non-Stiles” factor affecting both is the small temperature-induced
shift in lmax , interpreted by Ala-Laurila et al. (2002) to indicate
slight variation of Ea itself with temperature (# 0.5 kcal0mol in the
temperature range studied). Second, the slope of log S versus
log 10l spectra continues to increase even in the temperaturesensitive wavelength domain where Stiles’ theory would expect
straight lines. Lewis (1955) accommodated this by introducing an
extra parameter, the degrees of freedom of vibrational modes m.
This, however, will enable a certain degree of tradeoff between
parameters Ea and m when fitting D-spectra like those in Fig. 5. A
direct comparison of the Ea values obtained for different pigments
would require the assumption that m does not vary. Higher values
of Lewis’ m produce a gentler departure from baseline of the
D-spectra, hence larger apparent Ea .
Discussion
The effects of changing chromophore from A1 to A2
The general objective of this and two previous studies (Koskelainen et al., 2000; Ala-Laurila et al., 2002) is to elucidate the
relation between spectral absorbance and thermal stability of visual pigments. Barlow (1957) first hypothesized that the two be
physically coupled, so that red-shifting the spectrum would always
carry a cost in increased thermal noise. The basic idea is that both
properties reflect the same underlying entity, the energy barrier for
activation. A low barrier permits the pigment to be activated by
low-energy (“red”) photons, but also makes it susceptible to activation by thermal energy alone. As a universal law, the idea has
been refuted theoretically as well as experimentally (Matthews,
1984; Goldsmith, 1989; Barlow et al., 1993, Koskelainen et al.,
2000). Despite this, a significant empirical correlation exists between lmax and thermal activation rate F in the sample of rod
pigments where both properties have been measured (Firsov &
Govardovskii, 1990; Fyhrquist, 1999). This is true also for lmax
versus dark noise in different cone types (Rieke & Baylor, 2000).
The chromophore switch from A1 to A2 in the same opsin, used
by many species of fish and amphibians as a quick way of
red-shifting spectral sensitivity, offers an attractively simple model
for studying the correlation experimentally. Donner et al. (1990)
measured dark noise in porphyropsin and rhodopsin rods of the
bullfrog, finding that the ca. 25 nm red-shift of lmax caused by the
A1 r A2 switch is accompanied by a ca. 10-fold increase in
the rate of thermal pigment activations F. In the present work the
417
same pigment pair was characterized with respect to a third crucial
parameter, the minimum energy for photoactivation Ea . We find
that this parameter also differs between the two pigments in the
direction envisaged by the Barlow hypothesis ~Ea,A1 5 46.5 kcal0
mol is significantly larger than Ea,A2 5 44.2 kcal0mol, P , 0.05).
The ratio of the Ea values is close to the hypothesis’ prediction
based on the lmax ratio ~Ea,A2 0Ea,A1 ' 0.951 vs. lmax,A1 0lmax,A2 '
0.952). However, the F ratio predicted from these activation
energies is ca. 50 ~FA2 0FA1 5 exp ~~Ea,A1 2 Ea,A2 )0RT !), that is,
clearly larger than the ratio ;10 found by Donner et al. (1990).
Thus, Barlow’s (1957) simple hypothesis as such fails to account
quantitatively for the difference in thermal stability even in pigments that differ only in the chromophore group. This is hardly
surprising, as it has long appeared that the energy barriers for
thermal and photic activation are different (Baylor et al., 1980;
Barlow et al., 1993; Firsov et al., 2002). The molecular mechanisms underlying the correlation between lmax and F must be more
complex.
The opsin of bullfrog rods
Our conclusions about the effect of the chromophore in bullfrog
rods are based on the assumption that the rhodopsin and the
porphyropsin are both coupled to the same opsin (cDNA sequence
published by Kayada et al., 1995). Several examples are known
where the same photoreceptor cell can express more than one
(even three) different opsins, either simultaneously or sequentially
through different states of development (Shand et al., 1988; Wood
& Partridge, 1993; Röhlich et al., 1994; Makino & Dodd, 1996).
However, no case has been reported where the chromophore
change would be linked to an opsin change. Fong et al. (1985) used
isoelectric focusing to study varieties of rhodopsin in the leopard
frog, Rana pipiens and the bullfrog, Rana catesbeiana. Although
the frog rhodopsins, in contrast to bovine rhodopsin, separated into
two forms, the authors did not find any differences between the
porphyropsin and rhodopsin fields of the adult bullfrog retina, nor
any differences associated with the A2 to A1 transition during
metamorphosis. Moreover, the focusing pattern of bullfrog opsin
was identical to that of the leopard frog, which has only A1 in the
adult retina. Thus, we think that the possibility of an opsin difference between the porphyropsin and the rhodopsin of the bullfrog
can be neglected.
The purity of pigments and spectra
Errors in the determination of Ea might arise if the spectra of other
photoreceptors, not only those of bullfrog “porphyropsin” rods, in
fact reflected some degree of A10A2 mixture. We looked for
possible contaminations by fitting in each case the “warm” absorbance spectra with A1 and A2 templates (Govardovskii et al.,
2000) and we estimate that admixtures larger than a few percent of
the other chromophore would have been detected. The bullfrog
rhodopsin spectrum at 28.58C was best fitted by a 100% A1
template with lmax 5 501.7 nm. This is in good agreement with the
earlier result (lmax 5 501.6 nm) obtained by Govardovskii et al.
(2000). The crucian carp spectra were best fitted by 100% A2
templates, with lmax 5 526.3 nm for the rod spectrum at 28.58C
and lmax 5 619 nm for the L-cone spectrum at 218C. Likewise, the
“purified” bullfrog A2 spectrum (at 28.58C) was best fitted by a
100% A2 template with lmax 5 525.2 nm. This value is somewhat
smaller than the value initially assumed for assessing the A10A2
absorbance ratio in the original mixed spectrum, 527 nm based on
418
Firsov et al. (1994). We would like to emphasize, however, that the
method is quite robust against small changes in assumed lmax ,
especially when only spectra with high A2 proportion are accepted
for analysis. The mixed spectra used here had A2 percentages of
60–70%. Then the “purifying” corrections are small in the crucial
long-wavelength range and the Ea estimates for the A2 pigment
would in fact change little even if no corrections were made.
A2 pigments compared with A1 pigments
The sample of Ea values for A2 pigments now available allows two
conclusions. First, there is no unique and necessary connection
between lmax and Ea among these any more than among A1
pigments. Second, A2 pigments as a group have lower activation
energies than A1 pigments. The ranges of Ea values encountered in
the pigments hitherto studied by the same method are A2: 38.3–
44.2 ~n 5 5) and A1: 44.3–50 kcal0mol ~n 5 7) (Srebro, 1966;
Koskelainen et al., 2000; Ala-Laurila et al., 2002). Although the
highest A2 value (bullfrog porphyropsin) and the lowest A1 value
(Bufo marinus rhodopsin) virtually coincide at 44.2– 44.3 kcal0
mol, the ranges remain nonoverlapping.
Acknowledgments
We thank Mr. Antti Huotarinen for skillful technical assistance. This work
was supported by the Academy of Finland (grants 49947 and 36154) and
by the Finnish Graduate School of Molecular Nanotechnology.
References
Aho, A.-C., Donner, K., Hydén, C., Larsen, L.O. & Reuter, T. (1988).
Low retinal noise in animals with low body temperature allows high
visual sensitivity. Nature 334, 348–350.
Ala-Laurila, P., Saarinen, P., Albert, R., Koskelainen, A. & Donner, K. (2002). Temperature effects on spectral properties of red and
green rods in toad retina. Visual Neuroscience 19, 781–792.
Allen, D.M. & McFarland, W.N. (1973). The effect of temperature on
rhodopsin-porphyropsin ratios in a fish. Vision Research 13, 1303–1309.
Ashmore, J.F. & Falk, G. (1977). Dark noise in retinal bipolar cells and
stability of rhodopsin in rods. Nature 270, 69–71.
Autrum, H. (1943). Uber kleinste Reize bei Sinnesorganen. Biologisches
Zentralblatt 63, 209–236.
Barlow, H.B. (1956). Retinal noise and absolute threshold. Journal of the
Optical Society of America 46, 634– 639.
Barlow, H.B. (1957). Purkinje shift and retinal noise. Nature 179, 255–256.
Barlow, R.B., Birge, R.R., Kaplan, E. & Tallent, J.R. (1993). On the
molecular origin of photoreceptor noise. Nature 366, 64– 66.
Baylor, D.A., Matthews, G. & Yau, K.-W. (1980). Two components of
electrical dark noise in toad retinal rod outer segments. Journal of
Physiology 309, 591– 621.
Bridges, C.D.B. (1956). The visual pigments of the rainbow trout (Salmo
irideus). Journal of Physiology 134, 620– 629.
Bridges, C.D.B. (1964). Effect of season and environment on the retinal
pigments of two fishes. Nature 203, 191–192.
Bridges, C.D.B. (1967). Spectroscopic properties of porphyropsins. Vision
Research 7, 349–369.
Bridges, C.D.B. (1972). The rhodopsin-porphyropsin visual system. In
Handbook of Sensory Physiology, VII/1. Photochemistry of Vision, ed.
Dartnall, H.J.A., pp. 417– 480. Berlin-Heidelberg-New York: Springer.
Bridges, C.D.B. & Yoshikami, S. (1970). The rhodopsin-porphyropsin
system in freshwater fishes—1. Effects of age and photic environment.
Vision Research 10, 1315–1332.
Dartnall, H.J.A. (1955). Visual pigments of the bleak (Alburnus lucidus). Journal of Physiology 128, 131–156.
Dartnall, H.J.A. (1972). Photosensitivity. In Handbook of Sensory Physiology, Vol. VII/1. Photochemistry of Vision, ed. Dartnall, H.J.A.
pp. 122–145. Berlin-Heidelberg-New York: Springer.
P. Ala-Laurila et al.
Dartnall, H.J.A. & Lythgoe, J.N. (1965). The spectral clustering of
visual pigments. Vision Research 5, 81–100.
Dartnall, H.J.A., Kander, M.R. & Munz, F.W. (1961). Periodic changes
in the visual pigment of a fish. In Progress in Photobiology, ed.
Christensen, B.C. & Buchmann, B., pp. 203–213. Amsterdam:
Elsevier.
Denton, E.J. & Pirenne, M.H. (1954). The visual sensitivity of the toad
Xenopus laevis. Journal of Physiology 125, 181–207.
de Vries, H. (1948). Der Einfluss der Temperatur des Auges auf die
spektrale Empfindlichkeitskurve. Experientia 4, 357–358.
Donner, K., Hemilä, S. & Koskelainen, A. (1988). Temperaturedependence of rod photoresponses from the aspartate-treated retina of
the frog (Rana temporaria). Acta Physiologica Scandinavica 134,
535–541.
Donner, K., Firsov, M.L. & Govardovskii, V.I. (1990). The frequency of
isomerization-like “dark” events in rhodopsin and porphyropsin rods of
the bull-frog retina. Journal of Physiology 428, 673– 692.
Firsov, M.L. & Govardovskii, V.I. (1990). Dark noise of visual pigments
with different absorption maxima. Sensornye Sistemy 4, 25–34. (In
Russian).
Firsov, M.L., Govardovskii, V.I. & Donner, K. (1994). Response univariance in bull-frog rods with two visual pigments. Vision Research
34, 839–847.
Firsov, M.L., Donner, K., & Govardovskii, V.I. (2002). pH and rate of
“dark” events in toad retinal rods: Test of a hypothesis on the molecular
origin of photoreceptor noise. Journal of Physiology 539, 837–846.
Fong, S.-L., Landers, R.A. & Bridges, C.D.B. (1985). Varieties of
rhodopsin in frog rod outer segment membranes: Analysis by isoelectric focusing. Vision Research 25, 1387–1397.
Fyhrquist, N. (1999). Spectral and thermal properties of amphibian
visual pigments related to molecular structure. Dissertationes Biocentri
Viikki Universitatis Helsingiensis 18099.
Goldsmith, T.H. (1989). Compound eyes and the world of vision research.
In Facets of Vision, ed. Stavenga, D.G. & Hardie, R.C., pp. 1–14.
Berlin-Heidelberg-New York: Springer-Verlag.
Govardovskii, V.I., Fyhrquist, N., Reuter, T., Kuzmin, D.G. & Donner, K. (2000). In search of the visual pigment template. Visual
Neuroscience 17, 509–528.
Hárosi, F.I. (1994). An analysis of two spectral properties of vertebrate
visual pigments. Vision Research 34, 1359–1367.
Kayada, S., Hisatomi, O. & Tokunaga, F. (1995). Cloning and expression of frog rhodopsin cDNA. Comparative Biochemistry and Physiology 110B, 599– 604.
Koskelainen, A., Hemilä, S. & Donner, K. (1994). Spectral sensitivities
of short- and long-wavelength sensitive cone mechanisms in the frog
retina. Acta Physiologica Scandinavica 152, 115–124.
Koskelainen, A., Ala-Laurila, P., Fyhrquist, N. & Donner, K. (2000).
Measurement of thermal contribution to photoreceptor sensitivity. Nature 403, 220–223.
Lewis, P.R. (1955). A theoretical interpretation of spectral sensitivity
curves at long wavelengths. Journal of Physiology 130, 45–52.
Makino, C.L. & Dodd, R.L. (1996). Multiple visual pigments in a
photoreceptor of the salamander retina. Journal of General Physiology
108, 27–34.
Makino, C.L., Groesbeek, M., Lugtenburg, J. & Baylor, D.A. (1999).
Spectral tuning in salamander visual pigments studied with dihydroretinal chromophores. Biophysical Journal 77, 1024–1035.
Marks, W.B. (1965). Visual pigments of single goldfish cones. Journal of
Physiology 178, 14–32.
Matthews, G. (1984). Dark noise in the outer segment membrane current
of green rod photoreceptors from toad retina. Journal of Physiology
349, 607– 618.
McFarland, W.N. & Allen, D.M. (1977). The effect of extrinsic factors
on two distinctive rhodopsin-porphyropsin systems. Canadian Journal
of Zoology 55, 1000–1009
Palacios, A.G., Varela, F.J., Srivastava, R. & Goldsmith, T.H. (1998).
Spectral sensitivity of cones in the goldfish, Carassius auratus. Vision
Research 38, 2135–2146.
Reuter, T. (1969). Visual pigments and ganglion cell activity in the retinae
of tadpoles and adult frogs (Rana temporaria L.). Acta Zoologica
Fennica 122, 1– 64.
Reuter, T.E., White, R.H. & Wald, G. (1971). Rhodopsin and porphyropsin fields in the adult bullfrog retina. Journal of General Physiology
58, 351–371.
Temperature effects on porphyropsin spectra
Rieke, F. & Baylor, D.A. (2000). Origin and functional impact of dark
noise in retinal cones. Neuron 26, 181–186.
Röhlich, P., van Veen, T. & Szél, A. (1994). Two different visual
pigments in one retinal cone cell. Neuron 13, 1159–1166.
Shand, J., Partridge, J.C., Archer, S.N., Potts, G.W. & Lythgoe, J.N.
(1988). Spectral absorbance changes in the violet0blue sensitive cones
of the juvenile pollack, Pollachius pollachius. Journal of Comparative
Physiology A 163, 699–703.
Srebro, R. (1966). A thermal component of excitation in the lateral eye of
Limulus. Journal of Physiology 187, 417– 425.
Stell, W.K. & Hárosi, F.I. (1976). Cone structure and visual pigment
content in the retina of the goldfish. Vision Research 16, 647– 657.
St. George, R.C.C. (1952). The interplay of light and heat in bleaching
rhodopsin. Journal of General Physiology 35, 495–517.
Stiles, W.S. (1948). The physical interpretation of the spectral sensitivity
curve of the eye. In Transactions of the Optical Convention of the
419
Worshipful Company of Spectacle Makers, pp. 97–107. London: Spectacle Makers’ Co.
Tsin, A.T.C. & Beatty, D.D. (1980). Visual pigments and vitamins A in
the adult bullfrog. Experimental Eye Research 30, 143–153.
Williams, T.P. & Milby, S.E. (1968). The thermal decomposition of some
visual pigments. Vision Research 8, 359–367.
Wood, P. & Partridge, J.C. (1993). Opsin substitution induced in retinal
rods of the eel (Anguilla anguilla (L.)): A model for G-protein-linked
receptors. Proceedings of the Royal Society B (London) 254, 227–232.
Yoshizawa, T. (1972). The behaviour of visual pigments at low temperatures. In Handbook of Sensory Physiology, VII/1. Photochemistry of
Vision, ed. Dartnall, H.J.A., pp. 147–179. Berlin–Heidelberg–New
York: Springer.
Yoshizawa, T. & Horiuchi, S. (1969). Intermediates in the photolytic
process of porphyropsin. Experimental Eye Research 8, 243–244.